Abstract Extracellular vesicles (EVs) are believed to be an important means of transporting RNA and other signaling molecules between cells. Most types of cells in the human body secrete EVs which are each likely to have distinct biological functions. Studying different types of EVs and their role in normal physiologic function and disease-related processes requires a reliable means of capturing these particles from readily accessible body fluids, like blood or urine. Our goal is to develop a new way to isolate specific types of EVs based on the molecules present on their surface. Our hypothesis is that we can use the presence of two or more specific molecules on the surface of select EVs secreted by a particular cell type. Once we isolate highly purified populations we can more easily identify and measure their molecular contents. We will develop a new scalable approach to identifying combinations of surface molecules present on EVs from particular cell types and use this to identify multi-marker ?surface signatures? for several types of cells known to secrete EVs directly into the bloodstream. We will develop a new method to isolate the EVs with each of these surface signatures. This approach, which will only capture the EVs having all of the targeted surface markers, should greatly improve the capture specificity, thus, addressing one of the main shortcomings of the existing EV isolation methods. We will verify the purity of the isolated EVs by measuring the presence of each of the surface-signature markers on every EV using recently developed multi-marker assays and a new high-sensitivity single-EV characterization instrument. Generating highly purified EVs from specific cell types will enable more targeted studies of EVs than those that are presently performed. For example, we will identify specific regulatory RNA molecules in the purified EVs by next generation sequencing and measure the distribution of these molecules at the single EV level using the new instrumentation and assay techniques. Lastly we will refine the isolation methods to enable high-throughput isolation of EVs from several commonly used biofluids and automate the single EV characterization instrumentation to enable large studies that presently would be either impossible or too time-consuming to be cost-effective.